On Wed, 7 Jan 2015 09:23:39, Jakub Jelinek wrote:
>
> But I really don't like the busy waiting. You essentially want something
> like pthread_barrier_wait that libtsan isn't aware of, right?
>
Yes.
> As tsan is only supported on x86_64-linux (and in the future could be on
> some other 64-bit linuxes), I wonder if we couldn't just simplify libgomp's
> config/linux/bar* for that purpose - libtsan doesn't intercept syscall(3)
> and doesn't intercept futex in particular, so we could have a busy waiting
> free synchronization primitive. Or another option is to just dlopen
> libpthread.so.0 and dlsym pthread_barrier_wait in it and use that, that way
> libtsan won't intercept it either.
>
That would be the honey pot. But unfortunately I can't reach it.
I tried this,
cat barrier.h
#include <dlfcn.h>
static pthread_barrier_t barrier;
static int (*barrier_wait)(pthread_barrier_t *);
__attribute__((constructor(101)))
void barrier_init()
{
pthread_barrier_init (&barrier, NULL, 2);
barrier_wait = (int (*)(pthread_barrier_t *))
dlsym (dlopen ("pthread.so.0", RTLD_NOW),
"pthread_barrier_wait");
}
But dlsym gives me only a pointer to the tsan-interceptor.
Bernd.
